MEASURING DYNAMIC PROPERTIES OF WIND TURBINE FOUNDATION SOIL IN RESONANT COLUMN - ISSUES AND CHALLENGES

Wind turbine foundations are subject to large dynamic wind loads. The horizontal wind loads create large shear forces and moments on the foundation, which may result in critical issues related to the stability and deformation of wind turbine generator (WTG) system. This study reviews and collects information about WTG foundations, soil dynamic properties, and foundation vibration theories for the dynamic analysis of WTG foundations. This study focuses on the measurement of dynamic soil properties of soil from a WTG site. Resonant column tests are used to establish shear modulus reduction and damping curves at small to medium strains. Procedures used in the laboratory testing are also described. The apparatus yielded reasonable and reliable test results that are consistent with available published results and models. Because of the heterogeneity of foundation soil under a large WTG foundation, the state of stress and strain history under WTG foundations are expected to be non-uniform; thus, three soil samples from different locations under one WTG turbine are dynamically characterized. Resonant column tests were conducted under different environmental conditions, and the effects of environmental parameters on soil dynamic properties are reported in this thesis. Resonant column testing allowed for the accurate measurement of shear modulus reduction and absolute shear modulus value of the soil specimens. Thus, with representative site soil, such as an undisturbed soil specimen, the resonant column will yield an accurate estimate of the shear modulus of the foundation soil at the field state of stress and strain history. Test results showed that shear modulus reduction curves of soils from different locations under the WTG foundation are significantly different. Soils from the south part of the WTG foundation showed a higher rate of reduction than soil from the north part of the WTG foundation. At the highest shear strain level, 3�10-5, recorded during one year of field observation (Yilmaz, 2014), the shear modulus reduction for three tested soils were 14%, 24% and 42%. In addition, the soil from the north part of the foundation had a higher absolute value of maximum shear modulus, G0 52 MPa while the shear modulus of soil from south was 42 MPa. The highest absolute shear modulus value of 88 MPa was from soil cuttings from advancement of the borings for the strain gauge placement. Strain history has a significant influence on soil dynamic properties. Three different loading paths were studied. Results indicate that large shear strain history temporarily decreases shear modulus at all strain levels but increases the low strain shear modulus (G0) in the long term. Comparison of virgin(intact soil specimen) and large strain experienced shear modulus reduction curves inferred that the soil is able to record the largest strain in history. The number of dynamic loading cycles reduced the shear modulus of the soil slightly, which is about 2 MPa, but it has no overall effect on shear modulus reduction curves. An increase in confining pressure resulted in an increase of shear modulus at small to medium shear strain level; however, confining pressure had a very limited effect on normalized shear modulus reduction curves for the tested foundation soils. Linear threshold shear strain tends to decrease with increase in confining pressure and to increase with increase of deviator stresses. These results indicate that state of stress affects the threshold shear strain of the tested specimens. The study of some of the environmental parameters (confining pressure, state of stress, shear strain history etc.) helps us understand how dynamic soil properties change with design conditions and let us develop a more reliable shear modulus estimation. One year of real-time field strain gauge records (from Yilmaz (2014)) shows that the shear strain of foundation soil falls within the range (10-6 to 10-4) of shear modulus reduction curves that is built by resonant column. Shear modulus reduction curves measured in the lab may be combined with real site information to provide an accurate estimation of foundation response to wind actions or to calibrate vibration equations specifically suitable to the wind turbine foundation system. These results show that there may be large variations in the dynamic properties of soils within a single foundation site and these results may challenge the models used in the design of the dynamic foundation systems